Anybody that doesn't think the Nikon 14-24 is comfortably better than (but not perfect) the Canon 16-35 MkII really hasn't used one.
I've owned a Nikon 14-24 for five years, for architecture and interiors. I had read all the reports, etc, but the first time I used it I didn't find it that
sharp. Not in the way it was being described, anyway. I began to think I had made a horrible mistake. To my enormous relief, I found that it was gratifyingly sharp when focused manually using live view.
I have not had the opportunity of testing it against the 16-35 or Zeiss primes, but I have had the pleasure of comparing it with a Canon 5D2/5D3 with 24 TS-E Mk 2 and 17 TS-E. The 24 TS-E Mk 2 is the sharpest wide-angle lens I have ever tried. I gawp at its edge-to-edge perfection, even with large amounts of shift. Again, great care is required with focusing, but the nature of these lenses means one tends to use them on a tripod, and therefore use of live view is natural. The Nikon 14-24 is not in the same league, but it is head and shoulders above any zoom I have bought, borrowed, or hired. For instance it is sharper and straighter at 24MM than either the Nikon 24-70/2.8 (not difficult) or the Canon 24-70/2.8 Mk 2 (on a 5D3) although the Canon had less CA in the corners than my own 14-24.
I should add that I do not believe the 14-24 is unique, but that by keeping the focal length range quite tight, and all wide-angle, it suffers less compromise than one that must cross from wide to moderately long focal lengths. I would expect a 16-35 to be equally capable, or better.
Since my experience five years ago, I have noticed a proliferation in focus calibration tools and software. I also wonder how many people are disappointed after the first day or week with a new lens, and return lenses with perfectly good optics, because their AF is not spot-on. The same thing goes for assessments or comparisons. First impressions are not always the right ones ...